Allow the HTML module to use the lxml HTML parser when parsing html file includes

common/lib/xmodule/xmodule/html_module.py

@@ -1,4 +1,5 @@
 import logging
+from lxml import etree
 
 from xmodule.x_module import XModule
 from xmodule.raw_module import RawDescriptor
@@ -26,3 +27,8 @@ class HtmlDescriptor(RawDescriptor):
 
     js = {'coffee': [resource_string(__name__, 'js/module/html.coffee')]}
     js_module = 'HTML'
+
+    @classmethod
+    def file_to_xml(cls, file_object):
+        parser = etree.HTMLParser()
+        return etree.parse(file_object, parser).getroot()

common/lib/xmodule/xmodule/xml_module.py

@@ -90,6 +90,16 @@ class XmlDescriptor(XModuleDescriptor):
             if xml_object.get(attr) is not None:
                 del xml_object.attrib[attr]
 
+    @classmethod
+    def file_to_xml(cls, file_object):
+        """
+        Used when this module wants to parse a file object to xml
+        that will be converted to the definition.
+
+        Returns an lxml Element
+        """
+        return etree.parse(file_object).getroot()
+
     @classmethod
     def from_xml(cls, xml_data, system, org=None, course=None):
         """
@@ -127,7 +137,7 @@ class XmlDescriptor(XModuleDescriptor):
                 filepath = cls._format_filepath(xml_object.tag, filename)
                 with system.resources_fs.open(filepath) as file:
                     try:
-                        definition_xml = etree.parse(file).getroot()
+                        definition_xml = cls.file_to_xml(file)
                     except:
                         log.exception("Failed to parse xml in file %s" % filepath)
                         raise